Gigantum is an open-source platform for developing, executing, and sharing analysis and computations. It runs behind the scenes to manage configuration, collaboration and versioning while you work. It automates the creation of versioned and containerized code so that your work is accessible, reproducible and as transparent as possible.
Our goal is to make science and data science move faster. We want people to directly access and build on each other's work without all of the technical hassles.
Furthermore, we believe that people should be able to work locally yet still have easy access to the cloud. We created Gigantum to make it simple to move functioning code, data, and environment back and forth between all kinds of resources and people.
Gigantum is currently released with support for Python 2/3 and R running in JupyterLab. You can easily install packages with apt, pip and conda, as well as add Docker snippets for more customized packages. More languages and development environments are coming soon!
Gigantum bundles data, code and environment into an integrated repository called a Project.
They can be created from scratch, imported as a file, downloaded from the Gigantum Cloud, published with a single click, and permissioned for controlled sharing and collaboration. Each Project contains a granular history of changes to data, code, environment and executions through a visual record of figures and searchable text so that you can find and inspect results.
Analyses are separated into individual Projects. A search bar and filters means you don't have to dig through file folders to get what you want.
Gigantum provides the typical set of package managers like
apt, but it also goes further to let you use tools that require more complex or manual installations.
Using Docker snippets, a broad set of tools can be incorporated into a Project's environment. So, if only one person on your team knows how to install a complicated library or package, once it is installed in a Project, anyone else can easily access it on their local machine.
Each Project can be customized using package managers and Docker snippets, and this environment travels with the Project when you share it.
Gigantum captures and presents your work in a high resolution and easy to interpret activity feed that lets you search through results, code, data and environment changes. Changes to the data, code, and environment are automatically versioned and displayed.
Versions are represented in the activity feed through automatically extracted data such as figures, results, and code snippets. With a single click you can roll back to the exact version of data, code and environment that produced a given output.
The activity feed keeps a detailed timeline with user attribution, embedded images, views into code, and access to rollback.
Gigantum runs on Docker, so it works anywhere Docker does. Projects are never tied to a specific machine and can be used locally on Linux, Mac, and Windows. The platform allows easy interaction with cloud storage, and you can publish and sync with a single click.
At the moment, Gigantum Cloud only provides storage and sharing with collaborators but soon we will provide compute capabilities and a searchable repository of publicly available Projects.
The cloud view page showing Projects that have been shared by other users or stored in the cloud.